In most emerging markets, the vast majority are excluded from mainstream financial services and must have their credit needs met by high cost lenders, if they are met at all. Efforts to drive financial inclusion by developing financial infrastructure have largely fallen short. For example, traditional credit bureaus in most emerging markets contain data only on the already banked, and often times only negative data such as delinquencies and defaults. Consequently, traditional bureaus offer little assistance to the estimated 4.5 billion “credit invisibles” globally, and can effectively serve as a tool of financial exclusion by proliferating blacklists to lenders.
FARM identifies predictive, non-financial data assets (nodes) excluded from traditional credit bureaus, and links (spoke) them to a central server (hub). The primary data asset used by FARM is the data collected by agents/micro- entrepreneurs that is proprietary to eKutir. This includes: plot soil quality; crop history; whether plot is irrigated; inputs used (herbicide, pesticide, seeds, fertilizer); crop quality; size of plot; crop yields; crop revenue by cycle; whether or not crops are financed; amount of financing; repayment history for crop financing; and behavioral data about the farmer (age of account, frequency of payment, compliance with crop advice, growth in yield over pre-relationship period). In addition, FARM will supplement the proprietary data asset with third party data from public and private sources. Initially, third party data will be from government agencies and from the agricultural supply chain (producer groups, cooperatives, suppliers/distributers) and gradually move towards other data sets like weather, geo-spatial mapping, mobile/utility subscriptions, public API stack [e.g. India Stack]. Using machine learning, FARM will create a self- perpetuating, intelligent scoring system for small farmers to identify their credit worthiness and for financial institutions to underwrite credit.
